Do one-sided consumer agents actually change market structure?

Choudary argues that consumer-side agents like Muse and Instinct are not, by themselves, the "natural endpoint of the agent economy." He distinguishes "a new locus of aggregation" from "a new mechanism of coordination": the former "changes where the consumer enters the market," the latter "can change the very structure of the market." Today, he writes, "we have agentic participants but don't have agentic markets yet," because "most agents today represent only one actor in that relationship," producing "a lopsided agent economy" where "intelligence is accumulating on one side of many market relationships without an equivalent change in how the other side participates."

The mechanism he gives runs through paired examples. A traveller's agent can know she will leave a day earlier for a nonstop itinerary or "will spend more if a disruption can be handled flexibly," but "if airlines and hotels remain represented only by published fares, room categories and availability feeds," the agent "can optimize among pre-existing offers" but "cannot negotiate with the supplier to reconsider what they are willing to offer." Give the airline an agent too, Choudary writes, and it "could reason over inventory, yield targets, loyalty status... and respond with alternate offers," so "the transaction is no longer produced by a buyer agent choosing from a catalogue" but "through mutual negotiation between represented parties." He runs the same logic for procurement agents against suppliers' "catalogues, price lists and responses to predefined RFQs," and cites Indeed's Apply For Me (job-seeker side) and Sourcing Assistant (employer side) as a live case where representing both sides turns hiring from matching "two static documents" into "interaction between two sets of objectives and constraints" — noting Indeed "temporarily paused the automatic application mode of Apply For Me after employer feedback."

This sets up a precondition for Does model capability determine who wins in agent negotiations?: that paper's benchmark already assumes both buyer and seller are agentic and measures what happens once they are, while Choudary's claim is that most real markets haven't reached that state yet — his "lopsided agent economy" is the condition the benchmark's two-sided setup presupposes away. It also sharpens Why do consumers delay adopting agentic search?: both describe one-sided agentic search as still operating inside an unchanged market structure, which is what Choudary means by an agent improving "navigation of the existing market" rather than remaking it. It bears on Can AI seller agents shift buyer preferences through description edits? as an early instance of the seller side beginning to respond in kind, though tweaking a description to an AI reader is a thinner form of representation than the negotiating seller-side agent Choudary describes.

The excerpt is the first part of a series ("More on that follows") and offers no data, benchmark, or named market where two-sided agentic representation has actually occurred — the airline, procurement, and travel scenarios are illustrative, not observed. It does not establish how fast sellers, suppliers, or employers will build matching agents, or what happens to price and surplus once they do; it only argues that without that symmetry, agent adoption is reshaping access to markets rather than the markets themselves, which implies that claims of agents disrupting commerce should be read skeptically wherever only the buyer side has an agent.